The goal of this Phase II STTR program is to develop a high-throughput, ultra-sensitive accelerator mass spectrometer for the detection and quantification of tritium in labeled biological molecules. Accelerator mass spectrometry (AMS) is a highly selective means for detecting tritium that can achieve a measurement sensitivity over 1000 times greater than decay counting. Existing AMS instruments capable of measuring tritium are designed to also measure 14C and other higher mass isotopes. However, significant advantages accrue to an AMS system dedicated to the measurement of tritium. One advantage is greatly reduced instrument size and cost. A second, and perhaps even more important advantage, is high sample throughput. The low natural abundance of 3H, which is more than 1000 times lower than that of 14C, means that smaller absolute quantities and concentrations of tritium can be detected in labeled samples with equal measurement accuracy. Consequently, tritium AMS can be used in conjunction with very small volume, high-throughput microfluidic sample processing systems. A unique feature of the proposed system is integration of the AMS with interfaces that permit rapid, direct introduction of discrete samples as well as continuous-flow monitoring of chromatography. The resulting design lends itself to high-throughput applications that are inaccessible to conventional AMS approaches. In Phase I we demonstrated detection of 3H-labeled solution samples at very low energy using an existing dual-isotope biomedical AMS instrument. This work has allowed us to design an AMS instrument that is truly comparable in size and cost to other major laboratory analytical instruments, but that provides the unique capabilities of AMS for detection of extremely small quantities and concentrations of 3H-labeled compounds. In Phase II, we will design, fabricate and test a dedicated 3H-AMS instrument with sample introduction interfaces for both continuous sample injection via HPLC, as well as rapid introduction of discrete samples via microfluidic technologies. The resulting integrated systems will provide dramatic improvements in sample throughput, speed of analysis, and measurement sensitivity compared with presently available analytical instruments. This program is a collaboration between Newton Scientific, Inc., and the Biological Engineering Division at the Massachusetts Institute of Technology.